of Instability of Electrified Drops, etc. 75 
9. When an alcohol meniscus is examined under instantaneous 
illumination as a current first starts to flow from it (i.e. during the 
intermittent stage), some of the shapes that the meniscus is seen to 
‘assume are indicated by “a” in Fig. 3. The peaked form tapers 
down to a point of vanishing diameter. 
With voltages at which the surface shows a great deal of 
agitation when viewed in ordinary light, it is seen under instan- 
taneous illumination that large masses and threads of liquid are 
‘pulled out of the surface. Some of the forms which have been 
observed are shown in Fig. 3 by the drawings “6,” “c,” “d,” “e” 
fand “ f.” 
_ 10. After the current from the conical form of discharge with 
aleohol has been allowed to flow for some time, the plate under- 
neath the point becomes wet with a disc of liquid which has a 
diameter about equal to the distance between the point and the 
plate. From the amount of liquid thus carried over, it appeared 
‘possible that the whole current from the alcohol point might result 
‘from the charges carried by minute drops leaving the charged 
‘surface and impinging on the plate. 
__ Experiments to test this pot were made on the conical form - 
of discharge in the open air, by noting what effect a blast of air, 
blown at right angles to the electric field between the point and 
‘plate, would have upon the current received by the plate. The 
Yeceiving plate was divided into two parts, one of which was 
earthed directly and the other of which was earthed through a 
galvanometer. When the discharge point was placed directly over 
the division line between these two parts and a moderate blast of 
air from a foot-bellows was directed along the surface of the plate 
from the part connected to the galvanometer towards the second 
part, the current reaching the galvanometer was reduced to zero 
although the blast had no appreciable effect upon the discharging 
‘surface. When the point was moved so as to be over a portion of 
the plate at some distance from its edge, the same stream of air 
reduced the current flowing through the galvanometer but did not 
_ Under similar circumstances a current from a metal point, 
carried as it is by gaseous ions, would not have been appreciably 
vaffected by the stream of air which was used. It is evident, 
therefore, that all of the current in this case was carried by 
‘drops of liquid, much larger than ordinary ions. 
_ A rough estimate of the size of these drops may be made from 
the observation that with a current of 2000 electrostatic units the 
conical point discharged alcohol at the rate of 10~* cubic centimetres 
per second. The point had a radius of 0:0189 cm., was situated 
13 cms. above the plate and was maintained at a potential of 
8300 volts. The electric surface density -on the drops must be’ 
